The Farewell…

by ochiai808

Alas, the end of the quarter draws near, yet before we depart on our separate ways, a final post is in order.  As the saying goes, we learn something new everyday, and it is quite the poetic notion that we have been learning about learning itself.  Specifically, my posts have focused on the effects of instrumental conditioning on behavior in children.  As we step back to look at the bigger picture, I will remind you once again of the author and titles of the articles I have discussed in the order they have been presented.

Azrin et al.: Dry-Bed Training-Rapid Elimination of Childhood Enuresis

Richard et al.: The Use of Omission Training to Reduce Self-Injurious Behavior in a Retarded Child

Conyers et al.: A Comparison of Response Cost and Differential Reinforcement of Other Behavior to Reduce Disruptive Behavior in a Preschool Classroom

Wahler et al.: Mothers As Behavior Therapists for Their Own Children

Patterson & Anderson: Peers as Social Reinforcers

Bijou: Patterns of Reinforcement and Resistance to Extinction in Young Children

Hanley et al.: Influencing Preschoolers’ Free-Play Activity Preferences-An Evaluation of Satiation and Embedded Reinforcement

Magrab & Papadopoulou: The Effect of a Token Economy on Dietary Compliance for Children on Hemodialysis

Luerson et al.: Sticker Charts-A Method for Improving Adherence to Treatment of Chronic Diseases in Children

Billings et al.: Can Reinformcent Induce Children to Falsely Incriminate Themselves?

We first dove into the idea the discussion of whether instrumental reward or instrumental punishment was more effective in producing significant and robust behavioral results in children.  The articles by Azrin et al. and Richard et al. both revealed evidence that proved reward to be more advantageous.  In fact, their studies arose from the lack of efficiency of previous behavioral training methods which incorporated a type of punishment.  In the study by Azrin et al., the Urine-Alarm technique punished the child via a loud buzzer whenever the subject wet his or her bed.  In the study by Richard et al., a mild electric shock was administered whenever the child displayed self-injurious-behavior.  Both methods were not long lasting and took long to treat the undesired behavior.  The Dry-Bed technique, which focused on praise whenever the child correctly urinated and avoided wetting his or her bed, and the dispensing of a spoon of apple sauce whenever the child withheld from the self-injurious-behavior both proved to be more successful strategies.  Although in both studies there are many confounding variables, they support the idea that reward could be more effective than punishment in certain circumstances.

However, I do want to add that maybe the prior strategies just had low levels of efficiency and they both just so happened to be strategies implementing some type of punishment.  Therefore, any other strategy could have proved more effective, whether or not it was one with a rewarding nature.  The argument then could be that it was not a matter of reward versus punishment but was an argument between one specific strategy over another.  Because we are not able to test these type of experiments in completely controlled situations (for that is the nature of behavioral psychology on humans), it will be impossible to come to any final conclusions.  The reason why I bring this up is because the discussion on Conyers et al.’s article exhibited a case where a strategy implementing punishment seemed to be more successful than one implementing reward.  This shows that the efficiency of strategies depend on the context and the specific situations.  As I mentioned before, one method will not necessarily prove more effective than another across the board or in all types of situations.  In learning, there are no such things as golden rules.  Each circumstance is unique and therefore requires a type of learning tweaked to its own conditions.  In the case of Conyers et al., we have a group setting whereas in the previous two articles, the subject was being attended to individually.  Could this be the cause of the disparity?  In the studies by Azrin et al. and Richard et al., the subjects were never surrounded by anyone else with the behavioral problem.  Only they themselves were the ones being targeted for treatment.  However, in the article by Conyers et al., multiple children receiving treatment display disruptive behavior, and therefore, subjects are surrounded by others behaving similarly to them.  Subjects then may feel like their disruptive behavior is more “normal” and common than the subjects that wet their beds or participate in self-injurious-behavior.

Later on in the quarter I presented two other studies, that by Magrab & Papadopoulou and Luersen et al., which both focused on the effect of reinforcement (via sticker charts) on children suffering from chronic diseases.  These articles stemmed from the patients’ poor adherence to their medical treatment.  Although (in our debate on whether reinforcement or punishment is more effect) only reinforcement was tested, I believe that if they were to have tested a punishment type procedure, the punishment would be less effective than the reward.  This may be because, as I mentioned before, introducing a reward program gives the patients something to look forward to and want to work for.  It introduces something positive in their already adverse and tiresome life.  Medical difficulties, especially chronic diseases, not only burden the body but also have significant psychological ramifications.  It is not normal for a child to have to go through the demanding treatments, to be restricted from everyday activities, to be constantly worrying about their health.  Therefore, giving them some kind of encouragement, something rewarding, something positive is extremely important.  These cases are special from everyday conditions so, although it can be argued both ways for which is more effective (reinforcement vs. punishment), I feel that for clinical cases, especially those severe in nature, a reward system would prove triumphant.

As I mentioned before, there are many variables that cause a strategy to have varying levels of success.  Another aspect quite important in our discussion involves the therapists or instrumental reinforcers themselves.  For these articles, I focused on reinforcement for simplicity purposes and to not confound type of instrumental conditioning with the identity of the reinforcer.  The study by Wahler et al. focuses on the mother as being the therapist for her child where the study by Patterson & Anderson focuses on the child’s peers as the social reinforcers.  In both, the ultimate behavior of interest is the behavior of the child subjects.  However, the behavior that is being directly manipulated is that of the mother and the child’s peers.  Can such a strategy transfer a change in behavior from one individual to another?  And if so, can this prove to be more beneficial than directly changing the behavior of the subject?  These two studies produced data that supported the idea that not only is this type of indirect strategy successful, but it could be more beneficial such that it produces more long term results.  To discuss why this proved to be the case, I will remind you that Wahler et al. state that the parents’ “behaviors serve a large variety of stimulus functions” and “compose the most influential part of [the child’s] natural environment.” Therefore, the parents become the “source of eliciting stimuli and reinforcers which [produce] and [maintain] the child’s behavior” (Wahler, 114).  In addition, this type of stimuli provided by the parents constantly surrounds the child.  If a strategy works in a context that is unnatural for a subject, the resulting change of behavior may not generalize to other contexts and therefore not be long term.  In the same way, children spend the majority of their time with their peers, especially as they grow older and become less dependent on parents.  This again supports my overlying theme that there are countless things influencing the behavior in children.  I stated in one of my previous posts that the importance of learning lies not only in the individual, but in everything that influences the individual as well.  Successful behavioral improvement procedures must take into account the whole picture.  Learning is not an isolated entity; it is neither straightforward nor simple.  Instead, learning is complex, multifaceted, and involves many intricacies that must be acknowledged.  With that being said, changing a behavior is a very complicated process in which many dimensions must be taken with serious thought.

Another dimension that varies greatly is the specific characteristics of the reinforcement istelf.  Via the studies by Bijou and Hanley et al., we looked at both the patterns of reinforcement and the presence of an embedded reinforcement.  First, we saw that varying patterns of reinforcement, also referred to as intermittent reinforcement, produced a desired behavior that was more resistant to extinction than was a behavior produced by continuous reinforcement.  I provided some hypotheses why this might be the case:

Hypothesis 1) Because an intermittent type of reinforcement does not reward the subject on every trial, the subject could just assume it is just one of those trials that he or she will not rewarded.  They realize that, to be rewarded, they have to endure these trials with no reward and therefore will not be able to tell that they’ve entered an extinction phase or that anything has changed.  However, if the continuously reinforced subjects are not rewarded, this is not “normal,” and they realize that something has changed.  The “surprise” factor is different.

Hypothesis 2) The intermittent type of reinforcement already establishes a moderate amount of frustration on the subjects.  They are a bit upset that they do not receive rewards all the time, but this negative feeling is minor.  The continuous type of reinforcement establishes no frustration on the other group of subjects since they are always rewarded.  Although once no reward is administered, both group A and group B will be greatly frustrated, the difference between the initial levels of frustration and the final levels of frustration will be different for group A and for group B.  Group A will feel a greater change in frustration while group B, already frustrated a bit, will feel a smaller change in frustration.

There are other hypotheses that I would like to mention at this moment that I did not include earlier.  These stem from both Professor Blaisdell’s lectures and the Domjan readings.

Hypothesis 3) The children experiencing a partial reinforcement will be conditioned to continue to exhibit a behavior while feeling frustration.  The frustration thus turns into the stimulus.  Therefore, during the extinction period, the ever-present frustration elicits the continuation of behavior until the subject’s frustration builds up to cause the realization that reward no longer will follow.

Hypothesis 4) The children experience a reward following a nonreinforced trial (since it is a partially reinforced schedule).  Therefore, because subjects can remember if trials are reinforced or not (a phenomenon proposed by Eric Capaldi), the nonreinforced trial becomes a cue to continue behavior motivates responding in the hopes of receiving a reinforcement.  Similarly to hypothesis 3, this will continue until the exceedingly high level of frustration causes extinction.

The second characteristic of the reinforcement we discussed was that focused on in the study by Hanley et al.  We saw that the reinforcement does not have to be directly related to the behavior but may reside in its vicinity via embedded reinforcement.  We discussed embedded reinforcement as the reinforcement that attracted the children to the areas in which an increase in participation was desired.  Therefore, instead of reinforcing the behavioral participation in these zones, the zones themselves were reinforced.  Subject’s received the reward just by being in these areas whether or not they participated in the related activity.  This lead to the discussion of a dependence on the probability that once they are in the zones, they will subsequently interact with the activities surrounding them.  The data Hanley et al. found proved that this proved to be sufficient enough to change the child’s behavioral preference of the zones and participation in their respective activities.  I argued that this proves that reinforcement is such a powerful means that it can even indirectly alter behavior.  I would now like to propose that this embedded reinforcement may sometimes prove to be more beneficial in some situations.  This type of behavioral manipulation allows for the children to make their own decisions in regards to their behavior without direct manipulation on that particular behavior.  Therefore, they are able to choose how they want to behave without the expectation of a reward.  Although they are rewarded for being in the vicinity of the activity, they are never rewarded for actually partaking in the activity itself.  In a real life setting, reinforcement will not always follow a behavior so training the children not to expect a reward due to the behavior itself can allow the behavior to become more generalized.  In other words, subject’s have a choice in the experimental phase (in which the choice to participate in the behavior is not reinforced, just as it would be in an everyday situation) and still choose to partake in a behavior.  This can then provide a more accurate setting of the normal situations faced after experimentation trials.

Before I conclude my final post, I would like to discuss how powerful reinforcement can be.  We know that it can change behavior; however, we’ve only seen it change behavior into something more positive.  The outcome has always been beneficial for the subject and those surrounding individuals.  Therefore, the fact that behavior is going from negative to positive can be seen as a sort of push in the direction of a successful outcome.  It provides a type of motivation towards the desired result.  An argument can be made, then, that the opposite is true too.  If the outcome is objectionable and undesireable, this can lead to preexisting resistance against a change in behavior.  Therefore, all types of instrumental procedures will face a preliminary obstacle.  Would it then be harder for instrumental reward to create a result that is not wanted?  According to Billings et al., surprisingly, instrumental conditioning is so powerful that it can even produce outcomes that may be harmful to the subject.  He showed how subtle instrumental rewards can cause children to falsely incriminate their own selves.  Even if they are aware of the consequences of being guilty, the reinforcement is so strong that it does not matter in the overall picture.  The children still admit to things they didn’t do.

In summary, I would like to reiterate the overarching theme and specific points we have learned about.  What we should take away from these blog posts is that instrumental conditioning is not something that is clear cut.  It is not something that can fit any mold.  Instead, it is quite complex and depends on a myriad of influences which include the type of conditioning (reinforcement vs. punishment), the persons involved, the directness of the conditioning (whose behavior is directly conditioned), who the reinforcer is, the pattern of conditioning, and the present situation of the individual, just to name a few.  Therefore, to produce a strategy that is successful, all these ingredients must be weighed heavily and must be taken into account seriously.  There is so much we do not know about instrumental conditioning because that, as we’ve seen, is the nature of the beast.  It is so complex that we probably will never know all that there is to know about it.  However, no one can argue against the fact that we are influenced with instrumental conditioning in our everyday lives, whether we are aware of it or not, and that it has made all of us who we are today.